Objective To explore the risk factors for carbapenem-resistant Acinetobacter baumannii (CRAB) infection in patients in intensive care unit (ICU), and to construct and validate a nomogram risk prediction model. Methods Clinical data of ICU patients in a tertiary first-class hospital from June 1, 2022 to May 31, 2025 were collected retrospectively, and randomly divided into a modeling group and a validation group at a 7∶3 ratio. Based on data from the modeling group, univariate analysis and Lasso regression were used to screen characteristic variables, multivariate logistic regression analysis was adopted to determine independent risk factors for CRAB infection in ICU patients, and a nomogram prediction model was established. Internal validation of the model was conducted with data from validation group, and model performance was evaluated by receiver operating characteristic (ROC) curve, calibration curve, and decision curve. Results Among the 2 256 ICU patients, 194 developed CRAB infection, with an infection rate of 8.60%. Multivariate logistic regression analysis showed that the use of glucocorticoids or immunosuppressants (OR=3.21, 95%CI: 1.55-6.65), hypoalbuminemia (OR=10.20, 95%CI: 5.12-20.34), combination therapy with antimicrobial agents≥2 types before CRAB infection (OR=2.55, 95%CI: 0.99-6.51), duration of ventilator use≥7 days (OR=4.09, 95%CI: 2.56-6.53), previous infection before CRAB infection (OR=6.31, 95%CI: 3.80-10.49), and use of carbapenem antibiotics before CRAB infection (OR=9.93, 95%CI: 3.61-27.31) were independent risk factors for CRAB infection in ICU patients. Based on the above 6 independent risk factors, a nomogram model for predicting CRAB infection in ICU patients was established. The internal validation results showed that the area under ROC curve (AUC) values of the modeling group and the validation group were 0.909 (95%CI: 0.883-0.936) and 0.901 (95%CI: 0.862-0.939), respectively, indicating that the model had high predictive accuracy. The Hosmer-Lemeshow test had P-values of 0.991 and 0.521, respectively, indicating a good fit of the model. Conclusion The risk prediction model for CRAB infection in ICU patients constructed in this study has good discrimination and calibration, which can assist in identifying high-risk patients with CRAB infection and provide early warning references for accurate prevention and control.